DisasterEye
收藏资源简介:
DisasterEye是一个由穆罕默德·本·扎耶德人工智能大学创建的自定义数据集,旨在提供真实世界的灾难场景图像。该数据集包含2751张图像,分为8个类别,包括洪水、火灾、交通事故、地震后、泥石流、滑坡、正常和冲突。数据来源包括Google Images和YouTube等。数据集的创建过程涉及从多个来源收集图像,并将其分类为不同的灾难类型。该数据集的应用领域主要集中在无人机辅助的实时灾难检测,旨在通过优化的Transformer模型提高灾难检测的准确性和实时性,解决资源受限设备上的灾难管理问题。
DisasterEye is a custom dataset developed by Mohamed bin Zayed University of Artificial Intelligence, which is designed to provide real-world images of disaster scenarios. This dataset contains 2751 images, divided into 8 categories: flood, fire, traffic accident, post-earthquake, debris flow, landslide, normal, and conflict. The data sources include Google Images, YouTube and other platforms. The dataset creation process involves collecting images from multiple sources and classifying them into different disaster types. Its main application fields focus on UAV-assisted real-time disaster detection, aiming to improve the accuracy and real-time performance of disaster detection via optimized Transformer models, and address disaster management issues on resource-constrained devices.




